Re: [Xen-devel] [PATCH] xen/shim: stash RSDP address for ACPI driver

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 22.01.18 at 14:21, <wei.liu2@xxxxxxxxxx> wrote:
> On Mon, Jan 22, 2018 at 01:03:14PM +0000, Roger Pau Monné wrote:
>> On Mon, Jan 22, 2018 at 12:47:10PM +0000, Wei Liu wrote:
>> > --- a/xen/drivers/acpi/osl.c
>> > +++ b/xen/drivers/acpi/osl.c
>> > @@ -38,6 +38,10 @@
>> >  #include <xen/efi.h>
>> >  #include <xen/vmap.h>
>> >  
>> > +#ifdef CONFIG_PVH_GUEST
>> > +#include <asm/guest/pvh-boot.h>
>> > +#endif
>> > +
>> >  #define _COMPONENT                ACPI_OS_SERVICES
>> >  ACPI_MODULE_NAME("osl")
>> >  
>> > @@ -74,6 +78,11 @@ acpi_physical_address __init 
>> > acpi_os_get_root_pointer(void)
>> >                           "System description tables not found\n");
>> >                    return 0;
>> >            }
>> > +#ifdef CONFIG_PVH_GUEST
>> > +  } else if (pvh_boot) {
>> > +          ASSERT(pvh_rsdp_pa);
>> > +          return pvh_rsdp_pa;
>> > +#endif
>> >    } else if (IS_ENABLED(CONFIG_ACPI_LEGACY_TABLES_LOOKUP)) {
>> >            acpi_physical_address pa = 0;
>> 
>> Can this be done in a non-PVH specific way?
>> 
>> Can we have a global rsdp_hint variable or similar that would be used
>> here if set?
> 
> Who will be the anticipated user(s) other than PVH?

That's not so much the question here imo. Instead the issue I
see is that the way you code it it's really a layering violation.
Similar hackery was also rejected in Linux recently, iirc.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel